laryngeal quality of the Growl, were noticed in 17 of the 
deep hydrophone samples but were not dealt with further. 
The rest of the sounds tabulated were low-frequency 
sounds, made prominent aurally only by speedup of the 
playback transport. Those termed Low-Frequency Groans 
occurred only a limited number of times during a sample 
period, yet they showed great probability of occurring 
within sample periods during any time of the day, as shown 
in figure 11. The 20 c/s Short Pulses were also rather 
sporadic, but occurred in most samples listened to and 
during essentially all of the samples in the 2000 to 2200 
period. The 20 c/s Long Pulses are not included in these 
graphs because they were always detectable, even if very 
weak, within some part of each sample period during every 
hour of the day. The mean maximum S/N per sample (in a 
one-third octave band centered at 20 c/s) was 18.7 dB at 
the deep hydrophone and 23.4 dB at the shallow hydrophone. 
The Groans and Short Pulses occurred at both hydrophones 
about equally, and the curves represent presence of the 
sound in either or both hydrophones. 
Figure 12 shows the mean number of 20 c/s Long 
Pulses, Short Pulses, and Low-Frequency Groans per 
sample as a function of time of day. As in figures 10 and 
11 the curves here were plotted using the method of running 
averages. The number of Low-Frequency Groans increases 
sharply at 2000 at both hydrophones, while the peak num- 
ber of 20 c/s Short Pulses occurs at 2100. Around 0400, 
just before dawn, occurs a peak number of 20 e/s Long 
Pulses and what appear to be secondary peaks in the Short 
Pulses and Low-Frequency Groans. As explained earlier, 
the decision was made to include a short pulse well above 
20 c/s but below 100 c/s, not with the 20 c/s Short Pulses, 
but rather with the Groans (with which the questionable 
pulses were compatible in frequencies, but not in duration). 
The 20 c/s Pulses and the Low-Frequency Groans are 
obviously from powerful, distant sources (judging from the 
fact that the same pulses are heard on hydrophones 4 miles 
apart at approximately the same S/N). A correlation in 
31 
